Rapid state estimate for a class of dynamic models

Abstract

Even though well proved statistical methodologies in time series analysis can provide an optimal solution to well defined estimation and forecast problems, in many cases the optimal solution is not efficient from the point of view of computation. In many practical problems a shortcut to obtain good sub optimal solution is considered, for which it is possible to get rapid sub optimal solutions. Some consideration is made on how one can get to a simplified problem containing in fact the very kernel of the real problem getting rid of some complexity.